summaryrefslogtreecommitdiffhomepage
AgeCommit message (Collapse)Author
2018-07-14opentrack: this can be privateStanislaw Halik
2018-07-14opentrack: cleanup main_window::set_titleStanislaw Halik
2018-07-14compat/meta: fixesStanislaw Halik
2018-07-14fix buildStanislaw Halik
2018-07-14options/slider: remove constexprStanislaw Halik
it implies inline
2018-07-13appease analyzerStanislaw Halik
2018-07-11gui/init: sprinkle some `static'Stanislaw Halik
2018-07-11compat/macros: fix return typeStanislaw Halik
2018-07-11cmake: fix buildStanislaw Halik
2018-07-10fix buildStanislaw Halik
2018-07-10tidy: remove duplicatesStanislaw Halik
2018-07-10tracker/aruco: simplify abi testStanislaw Halik
The issue was that with LTO enabled, the `try_compile' call took way too long for a simple compile test. Rather, add a #define to Aruco to avoid the slow link. Bump aruco ABI version. Please update your opentrack/aruco and/or opentrack/opentrack-depends repos.
2018-07-10cmake/platform: cleanupStanislaw Halik
2018-07-10cmake: spew out effective flagsStanislaw Halik
2018-07-10fixup! cmake: renameStanislaw Halik
2018-07-10cmake/msvc: fix flags duplicating themselvesStanislaw Halik
2018-07-10tracker/tobii-eyex: working (?) prototypeStanislaw Halik
2018-07-10move to dot.githubStanislaw Halik
2018-07-09cmake: renameStanislaw Halik
2018-07-09tracker/tobii: add prototype support for accumulative modeStanislaw Halik
2018-07-09cleanup onlyStanislaw Halik
2018-07-08migration: actualy batch the updatesStanislaw Halik
2018-07-08options: fix 2 issuesStanislaw Halik
1. Calling valueChanged didn't invoke machinery in value<t>, only base_value aka value_. There's a fast path in value<t>::type() despite the pessimization. 2. Split global scope stuff into options::globals from the options::globals stuff 3. Adjust usages
2018-07-08modernize onlyStanislaw Halik
2018-07-08cmake/msvc: fix building some other projectStanislaw Halik
2018-07-08dot.clang-tidyStanislaw Halik
2018-07-06tracker/rift: euthanize old versionsStanislaw Halik
No one uses them according to #781
2018-07-06tracker/hatire: remove explicit workaround optinStanislaw Halik
Tested-by: @Ec0nomist cf. https://github.com/opentrack/opentrack/issues/327#issuecomment-402288929
2018-07-06clang-tidy: remove duplicate testStanislaw Halik
2018-07-03filter/kalman: don't need operator new overload anymoreStanislaw Halik
MSVC 15.0+ supports correct alignment
2018-07-03cleanupStanislaw Halik
2018-07-03compat/calibrator: fix typoStanislaw Halik
2018-07-03clang-tidy: get rid of nonsenseStanislaw Halik
2018-07-03compat/sleep: don't pull windows.h into the headerStanislaw Halik
2018-07-03logic/shortcuts: destructure binding; avoid captureStanislaw Halik
wrt capture: If a lambda-expression [...] captures a structured binding (explicitly or implicitly), the program is ill-formed. -- P0588R1, Core issue 2313 Also it's unnecessary to make the removed lambda.
2018-07-02add dot.clang-tidyStanislaw Halik
2018-07-02opentrack/main-window: use more aggregate initializationStanislaw Halik
This is new in C++17, cf. P0017R1.
2018-07-02nothing importantStanislaw Halik
2018-07-02compat/meta: try make usableStanislaw Halik
2018-06-26options/group: add commentStanislaw Halik
We'll be moving away from QSettings usage. The intent here is to: 1) get rid of this cruft 2) allow notifying all values with given name that their value changed in one of these values, i.e. `valueChanged()' without connector usage
2018-06-26spline: split up spline APIStanislaw Halik
2018-06-26spline: don't use cubicStanislaw Halik
Cubic shows incorrect values for control points close to each other.
2018-06-26options/value: add missing dereference operatorStanislaw Halik
2018-06-26options/slider: add missing `operator!='Stanislaw Halik
2018-06-26options/base-value: rename to `value_'Stanislaw Halik
2018-06-26filter/accela: get rid of pointless `progn'Stanislaw Halik
2018-06-26dinput: fix joystick POV hat typoStanislaw Halik
2018-06-26dinput: simplify di_tStanislaw Halik
2018-06-26modernize C++ syntaxStanislaw Halik
No visible changes (hopefully).
2018-06-26compat/macros: remove incorrect membarStanislaw Halik
The memory barrier impl for GNU was incorrect since it didn't prevent CPU reordering. It only prevented compiler reordering.